By Petko Dinev
Mobile video surveillance using several unmanned platforms communicating over a wireless network can be an effective technology to inform command-and-control activities in conventional open battlefields and congested urban environments.
During the past few years, development of unmanned aerial vehicles (UAVs) for battlefield reconnaissance has proceeded rapidly, with remotely piloted vehicles capable of carrying considerable remote-sensing payloads in Iraq, Afghanistan, and other theaters. More recently, autonomous UAVs able to carry out missions with minimal human guidance have come on line.
While these assets have already shown that this technology provides valuable information with reduced risk to human aviators, additional technology is already available to reach a much higher level of functionality. Mobile video surveillance (MVS) using intelligent networked platforms will revolutionize command and control technology to an even greater extent than the current generation of UAV platforms has done.
Future MVS systems will have flexible sensor systems consisting of high-resolution multispectral image sensors, built-in machine intelligence of a high order to distinguish threat targets from background objects, peer-to-peer wireless networks that allow swarms of fixed and mobile sensor assets to automatically coordinate their activities in advance of guidance from human commanders.
The MVS vision
Imagine a cadre of soldiers who are proficient with their weapons and are able to obey orders instantly, but have no initiative of their own. Such a force can be effective in a pitched battle, but hopelessly incompetent in a guerilla-style action. This corresponds to the current generation of UAVs.
Compare that force to a numerically inferior cadre of soldiers who have complete, up-to-the-minute information about the battlefield situation, the intelligence and training to use that information, and the capability of instant communications to coordinate their actions. Such a force would have superiority that would become more overwhelming as the battlefield environment became more difficult. Such is the advantage MVS technology offers over the present generation of unmanned assets.
MVS systems technology would operate at two levels: the individual asset level and the group (formation) level.
At the asset level, autonomous vehicles consist of a locomotion system, a navigation system, and a smart sensory system. UAVs are an example of an aerial locomotion system, but other types of locomotion systems are under development as well. While most UAVs still operate only under the immediate control of a remote pilot, UAVs with autonomous navigation systems are already in service. The MVS vision requires autonomous navigation.
Smart sensory systems include not only individual sensors, such as thermographs and CCD cameras, but also full-function digital computers capable of interpreting the information the sensors gather, making decisions based on that interpretation, and communicating decisions and information to other parts of the system. For example, a smart video- sensing system would be capable of seeing a moving target, determining whether it actually was a target, and providing navigation settings necessary to keep the target in view.
Consider how MVS assets might work at the formation level. Several smart assets work together as a distributed-intelligence network. They can work together (peer-to-peer) to make tactical decisions about how best to deploy themselves and what targets to concentrate on. They also all communicate directly with a command-and-control station that provides strategic tasking and can override tactical decisions made by consensus at the formation level.
It is important for hardware implementations of these systems to be as compact and light as possible. Unlike the combat situation, from a surveillance point of view there is nothing that a physically large asset can do that a small asset cannot do as well or better. For example, the current generation of large fixed-wing UAVs does an excellent job of observing from relatively high altitudes, but is at a distinct disadvantage at treetop level or below, or in urban settings. A much smaller rotorcraft-similar to a radio-controlled model helicopter-can get much closer to ground truth.
Available technology
Most of the components necessary to implement the MVS vision are available in commercial off-the-shelf (COTS) form. Smart cameras have been available for several years, and have grown in capability throughout that time. They are quite capable of recognizing and tracking objects, such as people, in real time and making rules-based decisions, such as whether they are wearing the right uniform. They have also demonstrated the ability to recognize suspicious activities, such as a person placing an object on a subway platform and leaving the area.
Most smart cameras also provide two-way digital communications as well as image outputs. Various standard interfaces, such as IEEE-1394 (FireWire), USB, and-most notably-Ethernet are conspicuously present.
Some smart cameras provide high performance in image-sensor terms as well. Multimegapixel cameras are available in smart form, as are high-resolution cameras operating at faster-then-video (faster than 30 frames per second) rates. Indeed, some smart cameras provide the flexibility of variable frame rate and resolution under software control. Thus, it is possible-and has been demonstrated for security applications-for such a camera to provide low-resolution video to a monitor just to show it is still alive and operating, and then automatically change settings to provide high-resolution images that allow positive target identification when needed.
By the same token, data communication over long-distance radio links (telemetry) is also a well-established technology. It is even possible to buy small-footprint COTS units that use satellite communications to establish beyond-line-of-sight data communications or use difficult-to-disrupt spread-spectrum techniques. Spread-spectrum techniques can use complex frequency-hopping codes to prevent eavesdropping and spoofing. Of course, for military applications the information itself would be encrypted as well, making for a very secure network.
What is needed to realize the MVS vision is to bring these various technological bits and pieces together into a standard package suitable for military use. Work must define the requirements, then select components fitting those requirements and build demonstration systems to ensure that they perform as needed under realistic conditions. There will, no doubt, be areas found along the way that need technical improvement. For example, better compression techniques may be needed to prevent network overloads. Perhaps expanded network bandwidths will be needed as well.
Beyond the technical development, however, we need military strategists thinking about how to best use this technology. We cannot wait until the technology is ready for military use to start incorporating it into strategic thinking. We need to know how we might use it in order to develop the requirements needed to guide the development effort.
Perhaps most important, however, we need to recognize that the MVS vision is valid-even imminent-and commit to making it an important part of our military-technology development effort.
Petko Dinev is president of Imperx Inc. in Boca Raton, Fla.
